Development Of Portable Spectrometer And Research Of Spectral Signal Recovery Method

As an important tool for physical and chemical analysis,spectrometer has been widely used.Traditional spectrometers are bulky and costly.Thanks to the continuous advancement of technology,portable spectrometers began to appear in the 1990 s.Such instruments are small in size,low in cost,and easy to use.However,due to the cost and other hardware conditions,their resolution is far from traditional instruments.The signal recovery method is a signal processing method that can improve the resolution of the spectrometer to a certain extent without increasing the hardware cost of the instrument.Because the spectrum restoration algorithm needs the instrument response function as a priori information to complete,so before using the restoration algorithm to restore the spectral signal,you need to try to obtain the instrument response function of the spectrometer.First,a portable spectrometer prototype was developed as an experimental platform for the spectral signal restoration method.The holographic concave grating was used as the core spectroscopic device of the portable instrument.The solidworks software was used to complete the design of the mechanical package of the instrument.The photosensitive SLA resin-based 3D printing technology printed the shell for the instrument.The test results show that the wavelength range measured by the instrument is about 300 nm to 800 nm from near ultraviolet to visible light.The spectral resolution of the instrument in the 300 nm to 690 nm band is about 1.1nm,and the resolution of the 690 nm to 800 nm band is about 1.8nm.Wavelength dependent.In order to establish the instrument response model of the spectrometer with the wavelength change,the pulse response signal generated by the low-pressure mercury argon gas discharge lamp was used as the excitation to measure the instrument response curve.For the shortcomings in the instrument response function model of the traditional pure Gaussian model spectrometer,A model of convolution of Gaussian and exponential functions is used to fit the measured instrument response curves of different wave bands respectively.Compared with the fitting results of traditional pure Gaussian theoretical models,it is found that the proposed model has higher fitting accuracy.The correspondence between the parameters of the convolution model and the wavelength is fitted using polynomials,thereby obtaining the instrument response function that varies with the wavelength.An improved signal restoration algorithm is proposed,which is combined with the instrument response function that changes with wavelength to form a complete spectral signal restoration method.Finally,in order to verify the effectiveness of the method in this paper,a signal restoration experiment was carried out on the spectrum of the low-pressure mercury-argon gas discharge lamp,and a part of the spectrum where aliasing occurred was compared and analyzed.The results show that the spectral restoration effect of the improved spectral signal restoration method proposed is better than that of the traditional method.